The worst specialist bowlers in Test cricket, and the worst team
Some readers have suggested that I should look at the worst bowlers in Test cricket the same way I have looked at the worst specialist batsmen

Martin Williamson
For the batsmen I had a very effective measure, the Batting Position Average, which could be used to identify a specialist batsman, in addition to other measures. We do not have such a measure for bowlers and we have to improvise.
Let me list down the criteria for selection.
1. The bowler must have played in a minimum of 15 Tests.
2. The bowler should have bowled, on an average, a minimum of 150 balls per Test. This excludes casual bowlers.
3. The bowling average should be above 40.00. Fair enough condition.
5. I have also excluded bowlers such as Mohammad Rafique, who have a bowling average between 40 and 50 and a difference in average values (bowling average minus batting average) less than 30. There is no way a quality player such as Rafique should get in this collection of incompetents.
This gets us a list of 18 bowlers.
It would be very simple to rank these based on the Bowling Average and that table would as well be enough. However I have done a simple additional analysis of the constituent measures to bring out the level of bowling. The following measures are used.
1. The Bowling Strike rate.
2. The Bowling RpO.
3. The number of wickets captured per Test.
It must be remembered that the Bowling Strike rate and RpO are the two components which form the Bowling Average and I have separated these to let the readers judge the lack of effectiveness.
The formula is given below.
Index=StrikeRate x 0.25 + RpO x 5 + (3.0 - W/T) x 10.The final table is given below.
Cty Bowler Mat Balls B/M Wkts Avge B/W RpO W/T Index
Then we have a couple of average Bangladeshi bowlers and a trio of average New Zealand bowlers.
The interesting entry then is Agarkar. How he could have played 26 Tests as an all-rounder is one of the mysteries of Indian cricket. I can understand his being selected for 191 ODI matches because he had one of the best strike rates as a bowler in ODIs (288 wickets in 191 matches). But 26 Tests, even conceding the Adelaide contribution, is inexplicable.
I am equally amazed that Mohammad Sami was selected for 33 Tests and captured fewer than 2.5 wickets per Test at a 50+ average. I will not make any comments except that Pakistan has had very competent and effective pace bowlers during this period and it is a surprise that Sami was on for such a long time.
Just to demonstrate the point that the Bowling Average is the most effective of all cricketing measures I have given below the table in decreasing order of Bowling average. Readers will note that there are very few significant changes.
Cty Bowler Mat Wkts BowAvge BatAvge
I am not able to select a wicket-keeper since keepers have, by definition, been excluded in the two selections. Readers can insert their own keeper. Khaled Mashud is one possibility. I will select 12 players so that the bowlers' selection can be effectively done between spinners and pace bowlers.
I will take the task seriously. Here is my team.
Bt/Bw Avge
Readers should remember that there is no malice in this selection and everything has been done in a lighter vein. Do not come in with serious objections. You are welcome to form your own teams.
Anantha Narayanan has written for ESPNcricinfo and CastrolCricket and worked with a number of companies on their cricket performance ratings-related systems